A method and system for determining a roll angle in an eye image is described. A center of an iris, an inner eye corner, and an outer eye corner in an eye image may be determined. The eye image may be normalized and transformed to a polar eye image using the center of the iris as a reference for the transformation. A shift associated with the inner eye corner and a shift associated with the outer eye corner may be determined. The shifts associated with the inner eye corner and the outer eye corner may be indicators of an orientation of the eye image or roll angle, and may be used for subsequent eye image processing and matching.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method comprising: obtaining an eye image; determining a location of a center of an iris in the eye image; generating a polar eye image based, in part, on the location of the center of the iris in the obtained eye image; determining, using one or more processors, locations of one or more of an inner eye corner and an outer eye corner in the polar eye image; determining one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image, the one or more role angles comprising a roll angle corresponding to a difference between a location of the center of the iris in the polar eye image and the location of one of the inner eye corner or the outer eye corner in the polar eye image; and storing, in a storage unit, the one or more role angles as biometric sample features.
The described computer method determines eye orientation (roll angle) from an image. It first finds the center of the iris in the eye image. Then, it converts the image into a polar representation using the iris center as the origin. It locates the inner and/or outer corners of the eye in this polar image. The roll angle is calculated as the difference between the iris center location and the location of the inner or outer eye corner in the polar image. This roll angle is then stored as a biometric feature.
2. The computer-implemented method of claim 1 , further comprising: determining a size of the iris in the eye image, the determining comprising: determining a contour of the upper eye lid and a contour of a lower eye lid in the eye image; determining the location of the center of the iris using the contours of the upper eye lid and the lower eye lid; and determining a number of pixels indicative of a diameter of the iris in the eye image.
The method described to determine eye orientation, as outlined in the previous description, also determines the size of the iris. This is done by finding the contours of the upper and lower eyelids in the original eye image. These contours are then used to locate the center of the iris. Finally, the diameter of the iris is measured in pixels.
3. The computer-implemented method of claim 1 , wherein generating the polar eye image based, in part, on the location of the center of the iris in the obtained eye image comprises: unwrapping the obtained eye image by converting pixel values associated with Cartesian coordinates to pixel values associated with polar coordinates; and generating a lateral orientation of the polar eye image.
The process of converting the eye image to a polar representation, as described in the method to determine eye orientation, involves unwrapping the image. This unwrapping is achieved by converting the pixel coordinates from Cartesian (x, y) to polar (radius, angle) coordinates, which generates a lateral orientation of the polar eye image, using the iris center as the origin.
4. The computer-implemented method of claim 1 , wherein determining the one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image comprises one or more of: determining a shift associated with the inner eye corner from a first predetermined point in the polar eye image; and determining a shift associated with the outer eye corner from a second predetermined point in the polar eye image, the second predetermined point being different than the first predetermined point.
Calculating the roll angle from the polar eye image, as described in the method to determine eye orientation, involves measuring shifts. This includes determining the shift or offset of the inner eye corner from a predefined reference point in the polar image and/or determining the shift of the outer eye corner from a *different* predefined reference point in the polar image. These shifts indicate the eye's orientation.
5. The computer-implemented method of claim 4 , wherein: the actions further comprise: determining an eye geometry feature according to a difference between a location of the inner eye corner in the polar eye image and a location of the outer eye corner in the polar eye image; and the second predetermined point is 180 degrees apart from the first predetermined point.
In addition to calculating shifts for roll angle determination as described in the previous claim, this method calculates an "eye geometry feature." This feature is the difference between the locations of the inner and outer eye corners in the polar image. Moreover, the reference point for the outer eye corner shift is set 180 degrees apart from the reference point used for the inner eye corner shift, in the polar image.
6. The computer-implemented method of claim 1 , further comprising: receiving a second eye image; aligning the second eye image with the first eye image using one or more role angles of the second eye image and the determined one or more role angles of the first eye image; and determining that the second eye image matches the first eye image when the second eye image is aligned with the first eye image.
The method described to determine eye orientation is further used for eye image matching. A second eye image is obtained and aligned with a first eye image using the roll angles calculated for both images. If, after alignment using these roll angles, the two images match, the system determines that the second eye image corresponds to the first.
7. The computer-implemented method of claim 1 , wherein determining, using one or more processors, the locations of the one or more of an inner eye corner and an outer eye corner in the polar eye image comprises: determining, before generating the polar eye image, locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image; and translating, from Cartesian coordinates to polar coordinates, the determined locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image.
The method of finding the inner and outer eye corners in the polar image, as described in the initial method to determine eye orientation, can be performed by first finding the eye corners in the original, non-polar eye image. Once the locations of the eye corners are known in Cartesian coordinates, these locations are translated into polar coordinates to find the corresponding locations in the polar image.
8. A non-transitory computer-readable storage medium comprising instructions, which, when executed by one or more computers, cause the one or more computers to perform actions comprising: obtaining an eye image; determining a location of a center of an iris in the eye image; generating a polar eye image based, in part, on the location of the center of the iris in the obtained eye image; determining locations of one or more of an inner eye corner and an outer eye corner in the polar eye image; determining one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image, the one or more role angles comprising a roll angle corresponding to a difference between a location of the center of the iris in the polar eye image and the location of one of the inner eye corner or the outer eye corner in the polar eye image; and storing, in a storage unit, the one or more role angles as biometric sample features.
A computer-readable storage medium stores instructions for determining eye orientation (roll angle) from an image. The instructions cause the computer to: obtain an eye image, find the center of the iris, convert the image to a polar representation using the iris center, locate the inner/outer eye corners in the polar image, and calculate the roll angle as the difference between the iris center and eye corner locations in the polar image. The roll angle is then stored as a biometric feature.
9. The non-transitory computer-readable storage medium of claim 8 , wherein the actions further comprise: determining a contour of the upper eye lid and a contour of a lower eye lid in the eye image; determining the location of the center of the iris using the contours of the upper eye lid and the lower eye lid; and determining a number of pixels indicative of a diameter of the iris in the eye image.
The computer-readable storage medium described to determine eye orientation, as outlined in the previous description, also includes instructions for determining the size of the iris. This is done by finding the contours of the upper and lower eyelids in the original eye image. These contours are then used to locate the center of the iris. Finally, the diameter of the iris is measured in pixels.
10. The non-transitory computer-readable storage medium of claim 8 , wherein generating the polar eye image based, in part, on the location of the center of the iris in the obtained eye image comprises: unwrapping the obtained eye image by converting pixel values associated with Cartesian coordinates to pixel values associated with polar coordinates; and generating a lateral orientation of the polar eye image.
The computer-readable storage medium for converting an eye image to a polar representation, as described in the method to determine eye orientation, involves instructions for unwrapping the image. This unwrapping is achieved by converting the pixel coordinates from Cartesian (x, y) to polar (radius, angle) coordinates, which generates a lateral orientation of the polar eye image, using the iris center as the origin.
11. The non-transitory computer-readable storage medium of claim 8 , wherein determining the one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image comprises one or more of: determining a shift associated with the inner eye corner from a first predetermined point in the polar eye image; and determining a shift associated with the outer eye corner from a second predetermined point in the polar eye image, the second predetermined point being different than the first predetermined point.
The computer-readable storage medium for calculating the roll angle from the polar eye image, as described in the method to determine eye orientation, involves measuring shifts. This includes determining the shift or offset of the inner eye corner from a predefined reference point in the polar image and/or determining the shift of the outer eye corner from a *different* predefined reference point in the polar image. These shifts indicate the eye's orientation.
12. The non-transitory computer-readable storage medium of claim 11 , wherein: the actions further comprise: determining an eye geometry feature according to a difference between a location of the inner eye corner in the polar eye image and a location of the outer eye corner in the polar eye image; and the second predetermined point is 180 degrees apart from the first predetermined point.
In addition to calculating shifts for roll angle determination as described in the previous claim, the computer-readable storage medium includes instructions to calculate an "eye geometry feature." This feature is the difference between the locations of the inner and outer eye corners in the polar image. Moreover, the reference point for the outer eye corner shift is set 180 degrees apart from the reference point used for the inner eye corner shift, in the polar image.
13. The non-transitory computer-readable storage medium of claim 8 , wherein the actions further comprise: receiving a second eye image; aligning the second eye image with the first eye image using one or more role angles of the second eye image and the one or more role angles of the first eye image; and determining that the second eye image matches the first eye image when the second eye image is aligned with the first eye image.
The computer-readable storage medium described to determine eye orientation is further used for eye image matching. It includes instructions to obtain a second eye image and align it with a first eye image using the roll angles calculated for both images. If, after alignment using these roll angles, the two images match, the system determines that the second eye image corresponds to the first.
14. The non-transitory computer-readable storage medium of claim 8 , wherein determining the locations of the one or more of an inner eye corner and an outer eye corner in the polar eye image comprises: determining, before generating the polar eye image, locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image; and translating, from Cartesian coordinates to polar coordinates, the determined locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image.
The computer-readable storage medium including instructions for finding the inner and outer eye corners in the polar image, as described in the initial method to determine eye orientation, can be performed by first finding the eye corners in the original, non-polar eye image. Once the locations of the eye corners are known in Cartesian coordinates, these locations are translated into polar coordinates to find the corresponding locations in the polar image.
15. A system comprising: one or more computers and one or more storage devices storing instructions that are operable and when executed by one or more computers, cause the one or more computers to perform actions comprising: obtaining an eye image; determining a location of a center of an iris in the eye image; generating a polar eye image based, in part, on the location of the center of the iris in the obtained eye image; determining locations of one or more of an inner eye corner and an outer eye corner in the polar eye image; determining one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image, the one or more role angles comprising a roll angle corresponding to a difference between a location of the center of the iris in the polar eye image and the location of one of the inner eye corner or the outer eye corner in the polar eye image; and storing, in a storage unit, the one or more role angles as biometric sample features.
The described system determines eye orientation (roll angle) from an image. It includes one or more computers and storage devices. These storage devices hold instructions that, when executed, cause the computer(s) to obtain an eye image, find the center of the iris, convert the image to a polar representation, locate the inner/outer eye corners in the polar image, and calculate the roll angle as the difference between the iris center and eye corner locations in the polar image. The roll angle is then stored as a biometric feature.
16. The system of claim 15 , wherein the actions further comprise: determining a contour of the upper eye lid and a contour of a lower eye lid in the eye image; determining the location of the center of the iris using the contours of the upper eye lid and the lower eye lid; and determining a number of pixels indicative of a diameter of the iris in the eye image.
The system described to determine eye orientation, as outlined in the previous description, also determines the size of the iris. This is done by finding the contours of the upper and lower eyelids in the original eye image. These contours are then used to locate the center of the iris. Finally, the diameter of the iris is measured in pixels.
17. The system of claim 15 , wherein generating the polar eye image based, in part, on the location of the center of the iris in the obtained eye image comprises: unwrapping the obtained eye image by converting pixel values associated with Cartesian coordinates to pixel values associated with polar coordinates; and generating a lateral orientation of the polar eye image.
The system for converting an eye image to a polar representation, as described in the method to determine eye orientation, involves unwrapping the image. This unwrapping is achieved by converting the pixel coordinates from Cartesian (x, y) to polar (radius, angle) coordinates, which generates a lateral orientation of the polar eye image, using the iris center as the origin.
18. The system of claim 15 , wherein determining the one or more role angles based on the determined locations of the one or more of the inner eye corner and the outer eye corner in the polar eye image comprises one or more of: determining a shift associated with the inner eye corner from a first predetermined point in the polar eye image; and determining a shift associated with the outer eye corner from a second predetermined point in the polar eye image, the second predetermined point being different than the first predetermined point.
The system for calculating the roll angle from the polar eye image, as described in the method to determine eye orientation, involves measuring shifts. This includes determining the shift or offset of the inner eye corner from a predefined reference point in the polar image and/or determining the shift of the outer eye corner from a *different* predefined reference point in the polar image. These shifts indicate the eye's orientation.
19. The system of claim 18 , wherein: determining the locations of the one or more of an inner eye corner and an outer eye corner in the polar eye image comprises: determining, before generating the polar eye image, locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image; and translating, from Cartesian coordinates to polar coordinates, the determined locations of the one or more of the inner eye corner and the outer eye corner in the obtained eye image; the actions further comprise: determining an eye geometry feature according to a difference between a location of the inner eye corner in the polar eye image and a location of the outer eye corner in the polar eye image; and the second predetermined point is 180 degrees apart from the first predetermined point.
In the eye orientation system, the inner and outer eye corners can be first found in the original eye image (before polar transform), then converted to polar coordinates. The system calculates an "eye geometry feature," which is the difference between the polar-image locations of the inner and outer eye corners. The reference point for measuring the outer eye corner's shift is 180 degrees from the inner eye corner reference in the polar image.
20. The system of claim 15 , wherein the actions further comprise: receiving a second eye image; aligning the second eye image with the first eye image using one or more role angles of the second eye image and the determined one or more role angles of the first eye image; and determining that the second eye image matches the first eye image when the second eye image is aligned with the first eye image.
The system described to determine eye orientation is further used for eye image matching. It receives a second eye image and aligns it with a first eye image using the roll angles calculated for both images. If, after alignment using these roll angles, the two images match, the system determines that the second eye image corresponds to the first.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
December 31, 2015
July 18, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.